OpenAI wants to go public at a valuation north of $1 trillion. It’s also currently losing money at a scale that would be alarming for almost any other company on the planet. Both of those things are true at the same time, and understanding how they fit together, rather than picking whichever one confirms what you already believed about AI, is the actual useful exercise here.
At an all-hands meeting on August 19, 2026, OpenAI CFO Sarah Friar told employees the company “will be a public company in 2027,” with the possibility of moving that timeline up if growth keeps accelerating. It was the first time staff had heard a concrete window from leadership, rather than vague signals about “eventually” going public. OpenAI had already confidentially filed its IPO paperwork with the SEC back on June 8, 2026, a standard move that lets a company begin the regulatory review process without disclosing financial details publicly until much closer to an actual listing.
The Valuation Math
OpenAI’s current private valuation sits at $852 billion, set in March 2026 when the company closed a $122 billion funding round, at the time the largest private funding round in history. The company and its bankers are reportedly targeting a public listing valuation above $1 trillion, and by at least one account, Sam Altman has called anything below that figure a “non-starter” for the eventual IPO.
To put that number in perspective against the company’s actual revenue: OpenAI’s annualized revenue run rate hit roughly $40 billion as of August 2026. At a $1 trillion valuation, that works out to a price-to-sales multiple of around 25 times trailing annualized revenue. That’s an aggressive multiple by almost any traditional standard, though “traditional standards” have arguably stopped applying cleanly to frontier AI companies for a while now. Whether a 25x multiple is justified depends almost entirely on how much you believe OpenAI’s growth trajectory, and its path to profitability, will hold up over the next several years, which is exactly the kind of thing an IPO prospectus is supposed to help investors evaluate.
The Growth Story Really Is Extraordinary
It’s worth taking the growth numbers seriously before getting to the less flattering part of the story, because they’re genuinely remarkable. OpenAI’s annual revenue went from roughly $3.7 billion in 2024 to over $40 billion in annualized run rate by August 2026, a trajectory that outpaces almost every major tech company’s growth curve at a comparable stage, including companies like Google, Meta, and Amazon during their own fastest-growth years.
That $40 billion figure came from a genuinely sharp inflection point. According to reporting citing people familiar with the matter, OpenAI’s revenue run rate had plateaued for roughly five months in 2026, holding close to $25 billion from around February through May, before accelerating sharply to $40 billion by August. Enterprise revenue reportedly drove most of that jump, surpassing consumer revenue for the first time in July, with the number of business customers growing 32% in a single month.
It’s worth being precise about what “$40 billion” actually means here, though, because it’s easy to read that figure as an audited annual revenue number, and it isn’t one. It’s an annualized run rate, meaning a single recent month’s revenue multiplied by twelve. It’s a genuinely useful management metric for tracking momentum, but it’s not the same thing as a GAAP-audited annual figure, and it can move around significantly month to month in a way that a full-year audited number wouldn’t.
The Losses Are Just as Real, and More Complicated Than One Number Suggests
Here’s where the story gets more nuanced than a single headline can capture. OpenAI’s most recently audited financials, covering fiscal year 2025, show $13.07 billion in booked revenue against a $20.92 billion operating loss. That’s the widely cited figure, and it’s a real, audited number, not a rumor or an estimate.
But there’s a separate, larger figure that’s also been reported: a $38.5 billion net loss for 2025. According to some reporting, roughly $30 billion of that net loss figure stemmed from a one-time, non-cash accounting adjustment tied to OpenAI’s corporate restructuring, not from ongoing operating expenses. Strip that one-time item out, and the underlying operating loss estimate drops closer to somewhere in the $8 to $21 billion range, depending on which specific costs different analysts choose to include or exclude. That’s still a very large loss for a company of OpenAI’s size and revenue. It’s just meaningfully different from “the company lost $38.5 billion running its actual business in one year,” which is a materially scarier-sounding claim that isn’t quite accurate.
A big piece of where that operating loss actually comes from: OpenAI paid Microsoft $17.2 billion for Azure compute access in 2025 alone, a figure that exceeds the company’s entire audited revenue for the year. Compute costs, not payroll or marketing, are the dominant expense driving these losses, and that dynamic isn’t expected to ease anytime soon. OpenAI’s own 2026 compute spending projections reportedly run as high as $50 billion for the year.
What OpenAI Is Projecting for the Future
OpenAI’s internal projections reportedly show the company reaching roughly $100 billion in annual revenue by 2029, with profitability arriving around the same timeframe, not before. Different analyses put the exact profitability year anywhere from 2029 to sometime in the early 2030s, and cumulative losses between now and that breakeven point have been estimated anywhere from roughly $44 billion to well over $100 billion across various leaked internal documents and analyst models, a range wide enough that it’s worth treating any single cumulative figure with real skepticism until it shows up in an actual audited filing.
The comparison analysts keep reaching for is Uber, which burned through somewhere in the range of $31 to $33 billion in cumulative losses before reaching profitability. OpenAI’s own internal projections suggest losses several times larger than that before it gets to the same point, for a product category that, unlike ride-hailing, doesn’t yet have fully proven unit economics at consumer scale. That doesn’t mean OpenAI won’t get there. It means the bet being made at a $1 trillion valuation is a genuinely large one, resting on OpenAI executing a scale-up that would be unprecedented even by the standards of the fastest-growing tech companies in history.
The Competitive Picture Is Messier Than the Headlines Suggest
Here’s a detail that tends to get lost in coverage focused purely on OpenAI’s own numbers: Anthropic, OpenAI’s closest direct competitor, reportedly posted an annualized revenue run rate of $65 billion in July 2026, well ahead of OpenAI’s $40 billion figure the following month. If that reporting holds up, it means the company most people still think of as the clear market leader in generative AI is currently being outpaced on revenue growth by its main rival, at least on this particular metric and this particular month.
That’s a genuinely useful check on any narrative that treats OpenAI’s dominance as settled or automatic. Anthropic has also reportedly filed confidentially for its own IPO, on June 1, 2026, at a reported valuation of $965 billion, putting it in almost the same valuation range as OpenAI despite the revenue gap running in Anthropic’s favor rather than OpenAI’s. CFO Sarah Friar has publicly acknowledged the possibility that Anthropic could go public before OpenAI does, reportedly telling staff that if Anthropic “pulls the cover off” its own confidential filing and lists first, that’s fine, framing it as OpenAI simply running its own race rather than treating a listing as a competition to win first.
Google’s Gemini adds a third variable that’s harder to pin down with a single clean number, since Gemini’s commercial revenue is bundled inside Alphabet’s much larger overall financials rather than reported as a standalone figure the way OpenAI’s and Anthropic’s numbers are. What’s clear directionally, though, is that this is not a two-horse race between OpenAI and everyone else. It’s a genuinely competitive field where the company with the most public name recognition isn’t necessarily the one growing fastest by revenue.
What This Means If You’re Trying to Read the Situation Clearly
A few things are worth holding at once here, rather than collapsing this into either “AI is a bubble” or “the growth justifies everything,” which are both simpler stories than what the numbers actually show:
- The growth is real and genuinely historic. Going from roughly $3.7 billion to $40 billion in annualized revenue in under two years is an extraordinary trajectory, regardless of how the profitability question eventually resolves.
- The losses are also real, but the specific number matters. A $20.9 billion audited operating loss and a $38.5 billion net loss driven partly by a one-time accounting item are different claims, and conflating them makes the situation look either better or worse than the underlying reality.
- A 25x revenue multiple is a bet on sustained, extraordinary growth continuing for years. That’s not impossible, but it’s a much larger assumption than most public market investors are typically asked to underwrite, which is likely part of why OpenAI’s own leadership has been cautious about locking in a specific listing date.
- OpenAI isn’t running away with the AI market by every metric. Anthropic’s reported revenue run rate has, by at least one recent snapshot, actually pulled ahead, which complicates any simple narrative about which company is “winning.”
None of this means OpenAI’s IPO won’t happen, or won’t succeed at a huge valuation. It might. But the gap between the growth story investors are being sold and the financial reality sitting underneath it is exactly the kind of thing that becomes unavoidable to scrutinize the moment a company actually goes public and starts filing quarterly reports instead of leaked internal projections. That scrutiny is coming either way. The only real question is how well the numbers hold up once it arrives.
Frequently Asked Questions
What valuation is OpenAI targeting for its IPO? OpenAI is reportedly targeting a public listing valuation above $1 trillion, up from its current private valuation of $852 billion, set in a $122 billion funding round in March 2026.
When will OpenAI go public? CFO Sarah Friar told employees in August 2026 that OpenAI “will be a public company in 2027,” with the possibility of an earlier debut if growth continues accelerating. The company confidentially filed IPO paperwork with the SEC in June 2026 but has not set a specific public listing date.
How much revenue does OpenAI actually generate? OpenAI’s annualized revenue run rate reached roughly $40 billion as of August 2026. Its most recently audited annual figure, for fiscal year 2025, showed $13.07 billion in booked revenue.
How much money is OpenAI losing? OpenAI’s audited 2025 financials show a $20.92 billion operating loss. A separately reported $38.5 billion net loss figure includes a large one-time, non-cash accounting adjustment tied to the company’s corporate restructuring and isn’t a direct measure of ongoing operating losses.
Is OpenAI still the clear leader in the AI industry by revenue? Not necessarily. Anthropic reportedly posted a $65 billion annualized revenue run rate in July 2026, ahead of OpenAI’s $40 billion figure the following month, suggesting the competitive picture is closer than public perception often assumes.
When does OpenAI expect to be profitable? OpenAI’s internal projections reportedly target around $100 billion in annual revenue and profitability by 2029, though various analyst estimates place actual breakeven anywhere from 2029 into the early 2030s.
